Reference: Candidate Sourcing Agent
referencerecruiting-sourcing
framework
Agent Framework
memory
Memory
retrieval
Retrieval / Vector Store
browser
Browser Automation
observability
Observability / Tracing
prompt_management
Prompt Management
deployment
Deployment Target
mcp_servers
MCP Servers
Architectural notes
Reference architecture for a recruiter's sourcing agent: takes a role spec, searches LinkedIn + GitHub + conference speaker lists, scores candidates, drafts personalized outreach. Why these picks: - Gemini for the long-context candidate-profile reasoning. - Zep for tracking outreach history per candidate over months. - Stagehand over raw Playwright because LinkedIn's DOM is unstable. - Greenhouse MCP for ATS round-trip. Common gotchas: - LinkedIn anti-bot defeats most naive scraping; budget for failures. - Personalization quality drops fast if research input is shallow.
Embed this stack
Show off your stack on your engineering blog or company page.
<iframe src="http://testserver/embed/stacks/recruiting-sourcing-agent" width="100%" height="380" frameborder="0" style="border: 1px solid #334155; border-radius: 8px;"></iframe>